View Full Version : epub creation tools


jbenny
10-31-2007, 04:21 PM
Considering the current lack of epub creation tools, other than InDesign or a text editor, I would like to get some input from everyone here.

In a tool designed specifically for epub creation, what features would you think most important? I'm not thinking about some WYSIWYG tool, like InDesign, but something for a more automatic and streamlined method of creating epubs, either singly or in bulk. Possibly building on existing Open Source tools.

Some basic markup capabilities could also be included. How do you think this could best be accomplished? What input formats do you think would be the most important to have, to accomodate a wide variety of already existing document formats (not just fiction)?

kovidgoyal
10-31-2007, 04:25 PM
I will be writing LRF->epub, txt(with lightweight markup)->epub, rtf->epub and of course plain html to epub.

The creation tools will have the ability to specify metadata and a table of contents and will of course attempt to automatically detect metadata and semantic information where possible.

These converters will leverage all the exisiting tools/infrastructure in libprs500

wallcraft
10-31-2007, 04:58 PM
On the level of "converters", I would like to see .opf to .epub. You get a .opf (OEB) e-book by exploding a .LIT e-book, or as the working files of MobiPocket Creator. This command line converter would essentially be mobigen (.html or .opf to .mobi), but for .epub (epubgen?).

A difference to mobigen, though, is that this would not necessarily be the end point of the process. Once you have an initial .epub document, then a .epub editor (once available) could be used to further refine the e-book.

kovidgoyal
10-31-2007, 05:03 PM
My html->epub converter will support opf (it already supports it for html->lrf conversion).

andym
10-31-2007, 05:35 PM
There are plenty of tools for converting to html, or for checking and validating the (x)html (and IIRC an epub validator is on the way). To me what would be most useful would be a tool that would producing the other package contents once you have a valid xhtml file.

jbenny
10-31-2007, 05:36 PM
I will be writing LRF->epub, txt(with lightweight markup)->epub, rtf->epub and of course plain html to epub.

The creation tools will have the ability to specify metadata and a table of contents and will of course attempt to automatically detect metadata and semantic information where possible.

These converters will leverage all the exisiting tools/infrastructure in libprs500

Great news. Always good to see more epub support available. What time frame are you looking at for a first version? If you need someone to help with testing and suggestions, I'll gladly help.

kovidgoyal
10-31-2007, 05:41 PM
Once my TODO stack shortens a little I'll start work on this, so a first version of lrf->epub should see the light of day in about 1-2 months, depending on how my Ph.D. is progressing :-)

jbenny
10-31-2007, 05:50 PM
There are plenty of tools for converting to html, or for checking and validating the (x)html (and IIRC an epub validator is on the way). To me what would be most useful would be a tool that would producing the other package contents once you have a valid xhtml file.

My thinking on this is: Not counting the XHTML content itself, there are only a handful of needed files which comprise an epub. These files all have a very well defined structure that is very much the same from one epub to another. This would make some type of templating system very possible.

Assuming you already had valid XHTML 1.1 content, filling in some form with certain meta information about the ebook (along with the reading order) would allow automatic creation of these other files and then the packaging into a Zip archive. Of course, the use of either an ISBN, a UUID or some other identifier would also have to be handled and perhaps generated.

BTW, you can break up the ebook into multiple chapters/sections and in fact, this is recommended. You could use one large XHTML file for the entire book, but this may not render (or page) as fast. If one file is used, you could still provide the navigation points (navpoints) needed by using bookmarks inside the document, instead of linking to separate files.

jbenny
10-31-2007, 05:53 PM
(and IIRC an epub validator is on the way)

Where did you hear about a validator? I have been suggesting this myself, but haven't heard any news about someone (IDPF maybe?) doing it.

RWood
10-31-2007, 05:58 PM
The extensions to the current libprs500 that kovidgoyal mentioned should more than meet my needs for epub publishing tools for the near future. I will keep a look out for it.

JSWolf
10-31-2007, 07:52 PM
I will be writing LRF->epub, txt(with lightweight markup)->epub, rtf->epub and of course plain html to epub.

The creation tools will have the ability to specify metadata and a table of contents and will of course attempt to automatically detect metadata and semantic information where possible.

These converters will leverage all the exisiting tools/infrastructure in libprs500
Can we get epub > LRF?

DaleDe
10-31-2007, 08:10 PM
Here is a link to a epub construction document.
http://www.hxa7241.org/articles/content/epub-guide_hxa7241_2007.html

Dale

kovidgoyal
10-31-2007, 08:31 PM
Sure i doubt epub to lrf will be feature complete, since LRF supports only a subset of the epub features, but basically yeah.

jbenny
10-31-2007, 08:31 PM
Here is a link to a epub construction document.
http://www.hxa7241.org/articles/content/epub-guide_hxa7241_2007.html

Dale

Yes, That article is very helpful. Another, more recent article can be found here: http://www.teleread.org/blog/?p=7442

JSWolf
10-31-2007, 09:55 PM
Of course, if the 505 does get epub support, we won't need epub to LRF.

RWood
10-31-2007, 10:54 PM
Of course, if the 505 does get epub support, we won't need epub to LRF.
But remember all of the people that may still have 500s if Sony does not update their firmware for epub.

andym
11-01-2007, 05:32 AM
My thinking on this is: Not counting the XHTML content itself, there are only a handful of needed files which comprise an epub. These files all have a very well defined structure that is very much the same from one epub to another. This would make some type of templating system very possible.

Assuming you already had valid XHTML 1.1 content, filling in some form with certain meta information about the ebook (along with the reading order) would allow automatic creation of these other files and then the packaging into a Zip archive. Of course, the use of either an ISBN, a UUID or some other identifier would also have to be handled and perhaps generated.

BTW, you can break up the ebook into multiple chapters/sections and in fact, this is recommended. You could use one large XHTML file for the entire book, but this may not render (or page) as fast. If one file is used, you could still provide the navigation points (navpoints) needed by using bookmarks inside the document, instead of linking to separate files.

Yep agree with all of that. I should have said 'xhtml file or files'.

As far as the validator is concerned I'm sure that was something Nick Whatsisname [sorry Nick - memory not what it used to be] mentioned in a post on here. could be wrong though,

kurt_da_nerd
06-16-2008, 05:26 PM
Im using InDesign to create ePub format books, and I've encountered a problem with the tabbing for the paragraphs. When creating a regular HTML ebook, you put in ^t^p. For ePub, the ^t^p doesn't seem to work. Does anyone know how to indent paragraphs for ePub? Please let me know.


thanks

Hadrien
06-16-2008, 07:37 PM
Yep agree with all of that. I should have said 'xhtml file or files'.

As far as the validator is concerned I'm sure that was something Nick Whatsisname [sorry Nick - memory not what it used to be] mentioned in a post on here. could be wrong though,

The validator name is epubcheck: http://code.google.com/p/epubcheck/

jotheman
06-18-2008, 08:26 AM
... but a little off-topic. :offtopic:

Viral marketing? Maybe.

MacZap
03-13-2009, 01:30 PM
someone ask for an epub validator? This is an online tool:
http://www.threepress.org/document/epub-validate/